online | memorias.ioc.fiocruz.br
In silico
analysis of amino acid variation in human respiratory
syncytial virus: insights into immunodiagnostics
Claudemir Souza1,2,3, Nilson IT Zanchin1,3, Marco A Krieger1,2,3, Adriana Ludwig1,2/+
1Fundação Oswaldo Cruz-Fiocruz, Instituto Carlos Chagas, Laboratório de Genômica Funcional, Curitiba, PR, Brasil 2Instituto de Biologia Molecular do Paraná, Curitiba, PR, Brasil
3Universidade Federal do Paraná, Programa de Pós-Graduação em Biologia Celular e Molecular, Curitiba, PR, Brasil
BACKGROUND The highly contagious nature of human respiratory syncytial virus (HRSV) and the gravity of its infection in newborns and vulnerable adults pose a serious public health problem. Thus, a rapid and sensitive diagnostic test for viral detection that can be implemented upon the first appearance of symptoms is needed. The genetic variation of the virus must be considered for immunodiagnostic purposes.
OBJECTIVES To analyse HRSV genetic variation and discuss the possible consequences for capture immunoassay development.
METHODS We performed a wide analysis of N, F and G protein variation based on the HRSV sequences currently available in the GenBank database. We also evaluated their similarity with homologous proteins from other viruses.
FINDINGS The mean amino acid divergences for the N, F, and G proteins between HRSV-A and HRSV-B were determined to be approximately 4%, 10% and 47%, respectively. Due to their high conservation, assays based on the full-length N and F proteins may not distinguish HRSV fromhuman metapneumovirus and other Mononegavirales viruses, and the full-length G protein would most likely produce false negative results due to its high divergence.
MAIN CONCLUSIONS We have identified specific regions in each of these three proteins that have higher potential to produce specific results, and their combined utilisation should be considered for immunoassay development.
Key words: HRSV - genetic variability - diagnostics - antigen detection
doi: 10.1590/0074-02760170013
Financial support: IBMP grants obtained from FINEP (grant nº 01.11.0286.00), BNDES (grant nº 11.2.1328.1).
CS was supported by fellowships from CAPES and CNPq. AL was supported by a fellowship from CNPq. NITZ and MAK are research fellows from CNPq. + Corresponding author: adriana.ludwig@pq.cnpq.br
Received 12 January 2017 Accepted 16 May 2017
Human respiratory syncytial virus (HRSV) is one of the main causes of lower respiratory infections, which result in severe burden, especially in children, the elder-ly and adults with chronic health problems. While there is not an updated epidemiological census, the incidence rate of HRSV infection was reported in a study by Nair et al. (2010). They estimated that 33.8 million children younger than five years old were infected with HRSV in 2005, and at least 3.4 million of these children were hospitalised due to severe complications.
HRSV is an enveloped virus belonging to the Pneu-moviridae family and the Mononegavirales order (Afonso et al. 2016) that has a negative sense single-stranded RNA genome encoding 11 proteins (Hacking & Hull 2002). The nucleocapsid (N), fusion (F) and attachment (G) proteins have been the main targets of therapeutic and diagnos-tic investigations (Buraphacheep et al. 1997, Terrosi et al. 2007, Green et al. 2015). The N protein binds strongly to viral RNA to form a helical structure (Buraphacheep et al. 1997). The F protein mediates fusion of the viral mem-brane with the host cell memmem-brane, delivers viral RNA to
the cytoplasm, and fuses infected cell membranes with those of healthy cells, which results in syncytia forma-tion (Hacking & Hull 2002). Together with the F protein, the HRSV G protein is responsible for binding host cell receptors and triggering infection (Melero et al. 1997). The G protein is a transmembrane glycoprotein that has an ectodomain containing two highly variable mucin-like regions (HVR1 and HVR2) that are also rich in glycosyl-ation sites and immune epitopes (Melero et al. 1997).
Antigenic studies using monoclonal antibodies against the F and G proteins led to the initial distinction between the HRSV groups A and B. Subsequently, this virus was classified into genotypes based on the nucleo-tide variation of the G protein in the HVR2 region. Thus far, 14 genotypes have been identified for the HRSV-A group and 25 have been identified for the HRSV-B group (Hu et al. 2017, Zheng et al. 2017).
Spe-1102 sequences for the N protein, 1426 for the F protein and 701 for the G protein. The sequence datasets were aligned using the MUSCLE algorithm of MEGA 7.0 software (Kumar et al. 2016).
The high number of retrieved sequences made it im-possible to generate a visually informative phylogeny tree. In addition, due the excessive redundancy of the da-tasets, we selected a defined number of representative se-quences to evaluate the variability of the HRSV proteins at the aa level. Initial phylogenetic trees for the N and F proteins that included all sequences retrieved from Gen-Bank were generated using the neighbour joining meth-od with a 500 replicate bootstrap test. Representative sequences of the major clades were randomly chosen to construct the final phylogenies and for further analyses. Sequences that presented high divergence, which were evident by long branches, were also chosen. Moreover, strain reference sequences available in the NCBI taxon-omy database (http://www.ncbi.nlm.nih.gov/taxontaxon-omy) were also included. Finally, sequences from other related viruses were used as the outgroup to root the trees. For the G protein, high diversity is represented by the vari-ous genotypes that have been described. Thus, we opted to use one representative sequence from each genotype (with a minimum of 70% aa sequence coverage). Due to the high divergence of the G protein and consequential low similarity and poor alignment with its homologues from other viral species, we opted to present an unrooted tree. The G protein alignment was obtained by PSI-Cof-fee, which was designed to align distantly related pro-teins (Kemena & Notredame 2009).
The final phylogeny for each protein was construct-ed by the maximum likelihood (ML) method using the aa substitution model indicated by the model selection. Bootstrap tests with 1,000 replicates for the N and F proteins and 500 replicates for the G protein were used to assess the reliability of the branches. Amino acid sequence divergences were obtained using p-distance, which calculates the number of aa differences per site between sequences, using the pairwise deletion option. All divergences and phylogenetic analyses were as-sessed using MEGA 7.0 software (Kumar et al. 2016).
HRSV similarity to other viruses - To evaluate the possible cross-reactivity of the HRSV in immunologi-cal tests, we investigated the similarity of the HRSV N, F and G proteins with their homologous proteins from other viruses that infect humans using NCBI BLASTP
and HRSV-B, respectively, were used due to the high divergence found between these groups. The protein se-quences corresponding to the best hits from each virus were aligned to the HRSV-A and HRSV-B sequences us-ing PSI-Coffee (Kemena & Notredame 2009).
RESULTS AND DISCUSSION
The N protein - N proteins are among the most abun-dant viral proteins of some Mononegavirales viruses, and they usually elicit a strong, and long-lasting humoral im-mune response in patients. N proteins are of special inter-est because they could be used for the development of sim-ple and rapid laboratory diagnostic assays for direct virus detection in clinical specimens (Petraityté-Burneikiené et al. 2011). These proteins also have strong species-specific patterns and can be good markers for early diagnosis (Li et al. 2015). The HRSV N protein is a 391-aa protein that homomultimerizes into a decameric ring that functions as a scaffold around which viral RNA associates to form nucleocapsids (Tawar et al. 2009, El Omari et al. 2011).
The HRSV N protein is highly conserved. It displays only a 1.89% overall mean aa divergence among the 1102 aa sequences analysed in this work. The maximum pair-wise divergence value is 5.38% between the AJZ69754 (isolate = “VN-817-8/10”, Vietnam) and AFX95846.1 (strain = “RSVA/GN435/11”, South Korea) sequenc-es. Alignment of the N protein sequences shows high conservation across the entire protein (Supplementary data, Fig. 1). The N protein phylogeny (Fig. 1A), which shows two major well-supported clades corresponding to HRSV-A and HRSV-B, revealed that the N protein is highly conserved as indicated by the short branch lengths. The mean divergence within the A and B groups is less than 1%, and the mean divergence between the groups is 4.3% (Fig. 1B).
To investigate the similarity of the HRSV N protein with its homologous counterparts from other human-infecting viruses, we performed serial NCBI BLASTP searches, which are summarised in Table I. Although this protein is highly conserved within HRSV, it shows significant similarity (61%) with only the N protein of the human metapneumovirus (HMPV), another Pneu-moviridae respiratory virus. Based on sequence com-parison, no significant hits were recovered for the HRSV N protein with other human-infecting Mononegavirales
viruses, although it has been predicted to show second-ary structure conservation (Barr et al. 1991).
from aa 160 to the C-terminal end (Supplementary data, Fig. 2), which could implicate in possible serological cross-reactivity. Consequently, polyclonal antibodies against this protein or monoclonal antibodies against these regions are not appropriate for HRSV diagnosis. In fact, cross-reactivity between the N proteins of these viral species has already been described for polyclonal antibodies and for monoclonal antibodies against these two mentioned regions (Zhang et al. 2015). Alternative-ly, monoclonal antibodies against the region between aas 30-160 could specifically identify HRSV since this re-gion is conserved only among HRSV sequences.
The F protein - The HRSV fusion glycoprotein F is a type I integral membrane protein that is synthesised as a 574-aa precursor, which is processed by a furin-like protease to excise a 27-residue glycosylated peptide (pep27), giving rise to the F2 (N-terminal region) and F1 (C-terminal region) polypeptides. These subunits are connected by two disulfide bonds and form a protomer that oligomerizes to form the mature trimeric F protein (Hacking & Hull 2002, McLellan et al. 2013). The pres-ence of these two furin sites separated by pep27 is a specific feature of the F protein from HRSV (Bolt et al. 2000), although the basic F protein characteristics are shared with other Pneumoviridae and Paramyxoviridae
family members. During cell entry, F glycoproteins un-dergo a conformational change that brings the viral and cellular membranes into proximity, ultimately leading to their fusion (Swanson et al. 2011).
A total of 1426 aa F protein sequences were ana- lysed. The overall mean aa divergence is 4.66%, and the maximum pairwise divergence value is 15.98% between the AHA61614.1 (isolate = “B6-9918”, Taiwan) and AHA61607.1 (isolate = “A10-6030”, Taiwan) sequences. Alignment of the selected F protein sequences is shown in Supplementary data, Fig. 3. The signal peptide pres-ents the highest divergence (mean of 35%), while the F1 and F2 subunits appear to be more conserved (diver-gence mean of 3.3% and 4.6%, respectively).
The F protein phylogeny, shown in Fig. 2A, was con-structed based on 22 selected HRSV sequences. As ex-pected, two well-supported groups are evident from the F protein tree, reflecting the separation of HRSV-A and HRSV-B. The mean divergence found within each group is nearly 4%, while the mean divergence between groups A and B is 10.1% (Fig. 1B). Phylogeny analysis revealed that two sequences that were deposited in GenBank as belonging to the HRSV-B group (AEN74947.1 and AHG54515.1) felt in the HRSV-A clade. Since HRSV classification is based mostly on the G protein, a possible explanation for this finding is the occurrence of recom-bination between co-circulating HRSV-A and HRSV-B. However, although recombination is common for some types of viruses, it is rare for HRSV, and it is most likely that these potential recombinants are due to either PCR or sequence assembly artifacts (Tan et al. 2012).
Although the HRSV F protein is less conserved than the HRSV N protein, it presents significant similarity with F proteins of a higher number of viruses (Table I). The F protein most similar to that of HRSV is from HMPV, sharing 36% identity and 59% similarity and length coverage of 88% of the aa sequence. Two segments of the F protein are more conserved in these viruses (Supple-mentary data, Fig. 4), including a region of 29 aas (posi-tions 36 to 64 of the HRSV F protein) that has 53% iden-tity and 84% similarity and a central region of 16 aas that has 81% identity and 100% similarity (positions 303 to 318 of the HRSV F protein). The most divergent regions are the signal peptide and the region from aas 100 to 130.
According to the F protein BLASTP scores, the vi-ruses identified as having similar F proteins are human parainfluenza virus 3, theNipah virus, the Hendra virus and themeasles virus. All of these viruses are from Para-mixoviridae, a related family of the Mononegavirales or-der (Afonso et al. 2016). However, the similarity and/or extensions of the BLAST alignment results are not high. The alignments are shown in Supplementary data, Figs 5-8 and illustrate only a few short regions that are well-conserved. No additional human-infecting viruses show
Fig. 1: analysis of the human respiratory syncytial virus (HRSV) N protein. (A) Maximum likelihood tree based on the JTT + I model. The boot-strap values are shown next to the branches (only greater than 50). The tree is drawn to scale, and the branch lengths are measured as the number of substitutions per site. The analysis involved 10 HRSV amino acid (aa) sequences, which are identified by their GenBank accession number
and NCBI subgroup (in bold). The N proteinof thebovine respiratory syncytial virus and human metapneumoviruswere used as outgroup.
since 1998 for preventing infection in children who have a substantial risk of developing severe forms of HRSV disease (Eiland 2009). The palivizumab epitope consists of a structural helix-loop-helix motif that has discontinu-ous residues within a 20-residue linear peptide (Corti et al. 2013, McLellan 2015). Among the Mononegavirales
human-infecting species, only the HMPV F protein is relevantly conserved in this region with the HRSV F pro-tein (Fig. 1C), as the linear sequence alignment shows 37% identity and 58% similarity. However, the possible antibody contacts with this type of discontinuous epitope make it difficult for reliable cross-reaction prediction.
The G protein - The HRSV G protein has a variable sequence length ranging from 282 to 321 aas depending on the genotype and has no sequence similarity with other Mononegavirales attachment proteins (Cui et al. 2013). It is highly glycosylated, containing 30-40 O-linked glycans and 4-5 N-O-linked glycans, which poten-tially constitutes up to 60% of the G protein molecular mass (Cui et al. 2013, McLellan et al. 2013).
The overall mean aa divergence among the 701 G protein sequences analysed is 27.6%, which is consistent with the expected low conservation. The maximum pair-wise divergence value is 61.4% between the AIF71060.1 (HRSV-B, isolate = “MI_B_55_12_12”, Italy) and P27021.1 (HRSV-A, strain RSB642) sequences. As mentioned above, the high divergence found for the G protein is represented by numerous distinct genotypes, with 14 genotypes assigned to HRSV-A (NA1-NA4, GA1-GA7, SAA1, CB-A, and ON1) and 25 genotypes for assigned to HRSV-B (BA1-BA12, BA-C, SAB1-SAB4, GB1-GB4, URU1-2, CB-B, CB1) (Hu et al. 2017, Zheng et al. 2017). The genotypes that had only short, partial sequences available (A: GA6 and SAA1; HRSV-B: BA6, BA8, BA11, SAB2 and URU1) were excluded from our analyses, and the phylogeny was inferred using 32 G protein sequences representing different genotypes (12 genotypes from HRSV-A and 20 from HRSV-B).
Fig. 3A shows the unrooted phylogeny of the HRSV G protein, which was constructed based on the entire protein. Presumably, two clades separate the HRSV-A and HRSV-B sequences, and the long branches suggest high divergence among the sequences. The relationships among the genotypes are not clear since the bootstrap values for most nodes are very low (< 50). This lack of resolution is not relevant to the main purpose on this work, and it may reflect the non-dichotomous nature of the HRSV lineage split. The mean divergence found
T A B L E I S u m m a ry o f B L A S T P r e su lt s f o r N
, F a
n
d G p
Fig. 2: analysis of the human respiratory syncytial virus (HRSV) F protein. (A) Maximum likelihood tree based on the JTT + G model. The bootstrap value is shown next to the branch (only greater than 50). The tree is drawn to scale, and the branch lengths are measured as the number of substitutions per site. This analysis involved 22 HRSV amino acid (aa) sequences, which are identified by their respective GenBank
acces-sion number and NCBI subgroup (in bold). The F protein from bovine respiratory syncytial virus and human metapneumoviruswere used as
outgroups. (B) Mean aa divergence (based on p-distance) found between the major clades HRSV-A and HRSV-B and the outgroup sequences from bovine respiratory syncytial virus (Bov RSV) and human metapneumovirus (Meta). The mean aa divergence found within the groups is shown in bold. (C) Amino acid sequence alignment of the palivizumab-specific antibody epitope (254-277 aa). One sequence from HRSV-A, HRSV-B, bovine respiratory syncytial virus and human metapneumovirus is represented in the alignment.
regions.
within the A and B groups is 11.1% and 6.34%, respec-tively, and the mean divergence between these groups is 47.8% (Fig. 3B), which are similar to the values de-scribed in previous studies (Melero et al. 1997).
Fig. 4 shows the alignment of the selected G protein sequences. The G protein can be divided into four regions that represent various levels of divergence: conserved region I, HVR1, a central conserved region and HVR2. A summary of aa divergence from these regions is pre-sented in Table II. The low variation of conserved region I and the central conserved region contrasts with the high variation of the remaining G protein regions. The central conserved region has 13 aas that are identical in all the sequences analysed. Between groups A and B, the mean divergence for HVR1 and HVR2 is 67.1% and 57.3%, re-spectively. Variation in these regions within group A is 14.4% and 18.4%, respectively. The sequence divergence is lower in group B, with a mean of 7.61% for HVR1 and 11% for HVR2. Differences in HVR1 are caused mainly by aa substitutions, while the differences in HVR2 are caused by both aa substitutions and the insertion and deletion of sequence segments (Cui et al. 2013). Thus, these regions are not suitable targets for prevention (Tan et al. 2013) and diagnosis since they could generate false negative results. In contrast, the 13-aa central conserved region would be an excellent target for diagnosis and prevention. This region (residues 164-176) is unglycosyl-ated and universally conserved among all clinical iso-lates (Maifeld et al. 2016). In fact, monoclonal antibodies against this region have been shown to successfully block HRSV infection and disease (Jorquera et al. 2016).
The high divergence of the HRSV G protein is also reflected in its divergence from other viruses. No signif-icant similarities were found to other human-infecting viruses (Table I). The absence of linear and structural similarity (Melero et al. 1997) of the HRSV G protein to other Mononegavirales attachment proteins probably prevents cross-reactivity.
Genetic variability and HRSV diagnostic tests - Available HRSV conventional diagnostic methods in-clude viral culture, viral antigens detection by direct or indirect immunofluorescent (IF) or enzyme-linked im-munosorbent assay (ELISA) and viral nucleic acid de-tection by reverse transcription-PCR (RT-PCR)
(Popow-Kraupp & Aberle 2011). Although these methods are useful, they require costly equipment and reagents and/ or trained operators and can be time consuming, restrict-ing their use to large centres since they are not available in all hospitals and cities. Additionally, clinical sample quality has a direct impact on the sensitivity and speci-ficity of these assays (Popow-Kraupp & Aberle 2011).
An ideal test for this disease should be rapid, inexpen-sive, easy to handle and compatible with use in remote areas that have no laboratory infrastructure. This can be achieved using immunochromatographic membrane assays designed with capture antibodies optimised for HRSV proteins, which could be the first test performed in acutely-infected patients, allowing for quick conten-tion measures to be taken. A second confirmatory test could be subsequently performed, preferable for geno-type identification and for treatment decisions.
The challenges of HRSV immunodiagnostics are re-lated to the false positive results caused by cross-reac-tivity and false negative results due to viral population variation or low-quality clinical samples. We analysed the HRSV proteins reported to be the most immunogen-ic in terms of their genetimmunogen-ic variation among the HRSV genotypes and their similarity to other respiratory vi-ruses and evaluated their potential use in or exclusion from diagnostic applications. It is important to note that the analyses described herein were performed using linear alignment of the aa sequences, which was meant to be a starting point for the development of diagnostic monoclonal antibodies and to be complemented by other methods, including further epitope mapping and three-dimensional structural protein analysis.
infection. Cross-reactivity could most likely occur with the N and F proteins of HMPV, a common, related re-spiratory pathogen found worldwide. This false positive result could be easily ruled out by a parallel G protein negative result, which could be interpreted as an infec-tion by other Mononegavirales viruses. Thus, combined use of these regions could provide the basis for improv-ing HRSV immunodiagnosis by diminishimprov-ing both false negative results and cross-reactivity.
AUTHORS’ CONTRIBUTION
CS performed the GenBank searches and amino acid re-covery; CS and AL carried out BLASTP, phylogenetic and di-vergence analyses; CS, AL and NITZ analysed the results and wrote the manuscript; MAK assisted in analysis of the results and manuscript writing; NITZ and MAK provided the funding and facilities for the study. All authors read and approved the final manuscript.
REFERENCES
Afonso CL, Amarasinghe GK, Bányai K, Bào Y, Basler CF, Bavari S, et al. Taxonomy of the order Mononegavirales: update 2016. Arch Virol. 2016; 161(8): 2351-60.
Barr J, Chambers P, Pringle CR, Easton AJ. Sequence of the major nucleocapsid protein gene of pneumonia virus of mice: sequence comparisons suggest structural homology between nucleocapsid proteins of pneumoviruses, paramyxoviruses, rhabdoviruses and filoviruses. J Gen Virol. 1991; 72(3): 677-85.
Bolt G, Pedersen LO, Birkeslund HH. Cleavage of the respiratory syn-cytial virus fusion protein is required for its surface expression: role of furin. Virus Res. 2000; 68(1): 25-33.
Brendish NJ, Schiff HF, Clark TW. Point-of-care testing for respira-tory viruses in adults: the current landscape and future potential. J Infect. 2015; 71(5): 501-10.
Buraphacheep W, Britt WJ, Sullender WM. Detection of antibodies to respiratory syncytial virus attachment and nucleocapsid proteins with recombinant baculovirus-expressed antigens. J Clin Micro-biol. 1997; 35(2): 354-7.
Cane P. Molecular epidemiology and evolution of RSV. In: Cane P, editor. Respiratory syncytial virus. Amsterdam: Elsevier; 2007. p. 89-114.
Corti D, Bianchi S, Vanzetta F, Minola A, Perez L, Agatic G, et al. Cross-neutralization of four paramyxoviruses by a human mono-clonal antibody. Nature. 2013; 501(7467): 439-43.
Cui G, Zhu R, Qian Y, Deng J, Zhao L, Sun Y, et al. Genetic variation in attachment glycoprotein genes of human respiratory syncytial virus subgroups a and B in children in recent five consecutive years. PLoS ONE. 2013; 8(9): e75020.
5(10): e008748.
Hacking D, Hull J. Respiratory syncytial virus - viral biology and the host response. J Infect. 2002; 45(1): 18-24.
Hu P, Zheng T, Chen J, Zhou T, Chen Y, Xu X, et al. Alternate circula-tion and genetic variacircula-tion of human respiratory syncytial virus genotypes in Chengdu, West China, 2009-2014. J Med Virol. 2017; 89(1): 32-40.
Jorquera PA, Anderson L, Tripp RA. Understanding respiratory syn-cytial virus (RSV) vaccine development and aspects of disease pathogenesis. Expert Rev Vaccines. 2016; 15(2): 173-87.
Kemena C, Notredame C. Upcoming challenges for multiple sequence alignment methods in the high-throughput era. Bioinformatics. 2009; 25(19): 2455-65.
Koo B, Eun C, Yoon T, Hoon J, Kyoung M, Sung H, et al. An isother-mal, label-free, and rapid one-step RNA amplification/detection assay for diagnosis of respiratory viral infections. Biosens Bio-electron. 2017; 90: 187-94.
Kumar S, Stecher G, Tamura K. MEGA7: molecular evolutionary genetics analysis version 7.0 for bigger datasets. Mol Biol Evol. 2016; 33(7): 1870-4.
Li K, Bai Z, Zhu H, Di B. Prospective evaluation of rapid antigen tests for diagnosis of respiratory viral pathogens. Transplant Proc. 2015; 47(6): 1790-5.
Maifeld SV, Bodrey R, Mok H, Chu M, Yu L, Yamagata R, et al. De-velopment of electrochemiluminescent serology assays to mea-sure the humoral response to antigens of respiratory syncytial virus. PLoS ONE. 2016; 11(4): e0153019.
McLellan JS, Ray WC, Peeples ME. Structure and function of respi-ratory syncytial virus surface glycoproteins. Curr Top Microbiol Immunol. 2013; 372: 83-104.
McLellan JS. Neutralizing epitopes on the respiratory syncytial virus fusion glycoprotein. Curr Opin Virol. 2015; 11: 70-5.
Melero JA, García-Barreno B, Martínez I, Pringle CR, Cane PA. An-tigenic structure, evolution and immunobiology of human respi-ratory syncytial virus attachment (G) protein. J Gen Virol. 1997; 78(10): 2411-8.
Nair H, Nokes DJ, Gessner BD, Dherani M, Madhi SA, Singleton RJ, et al. Global burden of acute lower respiratory infections due to respiratory syncytial virus in young children: a systematic re-view and meta-analysis. Lancet. 2010; 375(9725): 1545-55.
Petraityté-Burneikiené R, Nalivaiko K, Lasickienė R, Firantienė R, Emužytė R, Sasnauskas K, et al. Generation of recombinant
Popow-Kraupp T, Aberle JH. Diagnosis of respiratory syncytial virus infection. Open Microbiol J. 2011; 5: 128-34.
Swanson KA, Settembre EC, Shaw CA, Dey AK, Rappuoli R, Mandl CW, et al. Structural basis for immunization with postfusion re-spiratory syncytial virus fusion F glycoprotein (RSV F) to elicit high neutralizing antibody titers. Proc Natl Acad Sci USA. 2011; 108(23): 9619-24.
Tan L, Coenjaerts FE, Houspie L, Viveen MC, van Bleek GM, Wiertz EJ, et al. The comparative genomics of human respiratory syncy-tial virus subgroups A and B: genetic variability and molecular evolutionary dynamics. J Virol. 2013; 87(14): 8213-26.
Tan L, Lemey P, Houspie L, Viveen MC, Jansen NJ, van Loon AM, et al. Genetic variability among complete human respiratory syncy-tial virus subgroup A genomes: bridging molecular evolutionary dynamics and epidemiology. PLoS ONE. 2012; 7(12): e51439.
Tawar RG, Duquerroy S, Vonrhein C, Varela PF, Damier-Piolle L, Castagné N, et al. Crystal structure of a nucleocapsid-like nucleo-protein-RNA complex of respiratory syncytial virus. Science. 2009; 326(5957): 1279-83.
Terrosi C, Di Genova G, Savellini GG, Correale P, Blardi P, Cusi MG. Immunological characterization of respiratory syncytial virus N protein epitopes recognized by human cytotoxic T lymphocytes. Viral Immunol. 2007; 20(3): 399-406.
Zhang Y, Pohl J, Brooks WA, Erdman DD. Serologic cross-reactions be-tween nucleocapsid proteins of human respiratory syncytial virus and human metapneumovirus. J Clin Microbiol. 2015; 53(5): 1609-15.
